Gas purge required after property left empty

[FatGit]

Crucailly, the "forces" causing the pp's to equalise, aren't very strong. Propane doesn't even easily make it out of a boat!
If the partial pressure of the gas in the pipe is maintained at 1.020 bar, air doesn't make it in at all - or are you saying it does?
Not sure whose Law it would be...
.

They must surely (attempt to) equaliase beacuse of Brownian motion of the molecules.
Take the leaking pipe as being a semi-permeable membrane, and you can see that there will be a diffusion of gas out and air in.

 

[chrishutt]

Only for complete combustion to CO2 + H2O.
If there was less air the gas (methane) could still burn, but incompletely leading to the formation of CO.

There's always going to be ample air for complete combustion if all you have is a pilot light burning on its own.
The point is that ignition will not occur if the gas being delivered to the pilot light were too diluted with air. The critical point is probably the gas/air ratio at the point where the spark occurs. If that is normally say 8% then if the gas is prediluted by 50% this ratio becomes 4% which is below the lower limit of flammability.

 

[clf-gas]

But I don't know why. Only two burners, both with thermocouples which would have shut gas valves when the gas was turned off at the meter.
(J&S Warm air + circulator)

After 6 weeks the pilots wouldn't light, due, apparently to air in the pipe. How would it get there?

Bet namsag knows!

Some very interesting points raised and posted on this.

I would have thought, if the gas was issolated at the meter without pilots being extinquished then the pilot would burn until the FFD came into play or the gas supply pressure reach equilibrium with ambient pressure.

Now lets assume the FFD was passing or ther is a small leak on the pipework, sun shine warms room or falls directly on pipe and creates expansion of trace gas, this will omit from passing FFD/leak, room/pipe cools during evening causing contraction thus entraining air into pipe.
Now you would have a gas air mixture, this wont quite lite when holding a naked flame to the pilot due to low level presure/flow. though gas will be burning in very small quantities. Like if you light a gas fire without purging the air fully, you get that distintive hum.

Vodka has kicked in I am loosing track myself now.

 

[chrishutt]

Now you would have a gas air mixture, this wont quite lite when holding a naked flame to the pilot due to low level presure/flow.

The point is that ChrisR was only able to attempt ignition with the spark from the ignitors. For the reasons I explained above this might not work whereas a full flame (e.g. taper, gas lighter) might well have worked since it would have been applied to a larger area and would catch some gas where the air/gas ratio was within flammable limits, i.e. 5% - 15%.

 

[clf-gas]

Now you would have a gas air mixture, this wont quite lite when holding a naked flame to the pilot due to low level presure/flow.

The point is that ChrisR was only able to attempt ignition with the spark from the ignitors. For the reasons I explained above this might not work whereas a full flame (e.g. taper, gas lighter) might well have worked since it would have been applied to a larger area and would catch some gas where the air/gas ratio was within flammable limits, i.e. 5% - 15%.

I did read that Chris and i agree. The pilot injector omiting neat gas and entraining air should ignite, though a gas/air mixture entraining more air won't until gas/air reaches LEL at min.


Edited typo

 

[chrishutt]

With say 50% predilution with air the resulting pilot flame would be smaller than normal since the gas/air would reach the 15% - 5% range nearer to the burner outlet.

So an ignitor spark positioned to pass through the 15% - 5% gas/air ratio of a normal flame might then be outside of this zone with the smaller flame and so fail to ignite it.

Another problem with the smaller flame is that the velocity of the gas/air mixture would be higher nearer the injector and may even exceed the flame speed. This would make the flame unstable and prone to lift and extinguish.

 

[clf-gas]

This would make the flame unstable and prone to lift and extinguish.

Exactly. Over airation, then would only ignite with naked flame.

This is going away from Chris's original post as to why having to purge?.
Again I say due to FFd Passing or leak on installation pipework.

 

[chrishutt]

Thanks to TT we have an explanation for the ingress of air into the gas pipe. It only requires a very small rate of leakage, such as might well occur passing the seal of an FFD, over several weeks to make a significant difference to the proportion of gas in the pipe.

ChrisR couldn't see why it shouldn't ignite anyway. The answer is that it would have, if the ignition had been applied to the right part of the flame. But he was only able to use the existing spark ignitors that put the spark to far away from the tighter flame.

 

[clf-gas]

Then when ChrisR comes back off his hols he shall be the ENLIGHTENED one.

 

[chrishutt]

More likely he'll pull my explanation to pieces. We shall see.

 

[ChrisR]

Would I do that? Never!

No Eureka moments while I was away, but CH's comment about Flame Speed could be important here. Air is smaller molecules than methane (I expect!) so comes out of a small hole (pilot jet) much quicker.

I WAS eventually able to get a match to the pilot, by turning the whole burner assembly backwards, so it stuck out into the hall - I though I wrote that before, perhaps it was on the ARGI site.
To begin with there was no reaction, but as the gas concentration in the mix increased, the mixture did light from the match onwards, ie not from the jet, and it didn't stay alight when the match was removed.

SO it may be that the flame speed of the mixture (which would presumably vary with concentration), was too low at low concentrations, to light back to the pilot jet, with the weak mix coming out faster than 100% gas does.

I don't think we can assume anything about the small leak appearing like a semi-permeable membrane between liquids where osmosis happens. The intermolecular forces would be radically different. Molecules in liquids have forces holding them together, in gasses they bounce off each other. I think osmosis still partly applies to gasses but the membrane is not the same as a hole.

In a static situation, I don't think Brownian Motion would provide much of a force for the gas exchange, but the effects of the weather, as I and others suggested, could be quite significant.
Just gut feeling really.

A final thought - if there were TWO tiny leaks, it would be possible for one to let air in and the other if higher, gas out, which provides a force (gravity) for the gas exchange.
I'm feeling too middle aged right now to work out the pressure resulting from a 1 metre height difference - doing it mentally it's 0.05 mbar which isn't much, but it's there.

 

[clf-gas]

Air is smaller molecules than methane (I expect!) so comes out of a small hole (pilot jet) much quicker.

Dont know nowt about the molecular sizes of air/methane, though I would have thought given a pilot working at say, 4mb, then even if it was 90% air in pipework, would it not still be working at 4mb. assuming 20mb at the appliance, .

Flame speed, over airation (creating flame lift), so a naked flame in the correct position would show gas pressent but an ignition probe genrating a spark of 700+ degrees in the wrong position would'nt.

Thinking back, can remember a time when I had'nt purged fully, lit the radiant fire and it gave a loverly tune, was this air passing faster than neat gas (got me thinking emoticon)

 

[scatmanjohn]

Interestingly i was flicking through a CITB gas safety book and saw something that i had not heard of before but made me think of this post.

In a section on purging it listed all the usual reasons why you would purge.

But at the bottom it said " after depressurisation (Grahams law of diffusion)"

now i have not got a clue what that is even after googling it.

Could this be what could have happened in this case?

http://en.wikipedia.org/wiki/Graham's_law